Letters to the Editor
We thank Drs. Lawn and Wood for their interest in our article. To determine probability estimates for our decision analysis, we used all published data. We did not exclude data from developing world settings, but data from such settings were simply not available for most probability estimates at the time of the analysis. We agree that data from resource-poor settings should be used to assess the optimal timing of antiretroviral therapy (ART) initiation among HIV-infected patients with tuberculosis (TB) in such settings. This is supported by the following points.
The event rates that were most important in our decision analysis were mortality associated with HIV, immune reconstitution inflammatory syndrome (IRIS), and drug toxicity; treatment failure or relapse was also important.1 In the South African setting of Drs. Lawn and Wood, the HIV-associated mortality rate among patients with TB receiving ART is substantially higher than the rate used in our decision analysis, which was from a study conducted in North America.2 In addition, however, Lawn and Wood have recently reported on 19 TB/HIV-infected patients at their site who developed IRIS while receiving anti-TB therapy and ART, of whom 2 (10.5%) died.3 Of note, more than 75% of the 160 patients who received concomitant anti-TB therapy and ART started ART more than 2 months after initiating anti-TB therapy. There could have been even more severe IRIS cases had more patients initiated ART within 2 months of anti-TB therapy. It is also important to note that this IRIS-related mortality rate contrasts with a recent review of all reported TB-related IRIS cases (most of which were from developed countries), in which there were no deaths.4 An IRIS-related mortality rate of 10.5% would exceed the threshold value from our analysis (4.6%), and therefore favor deferred ART initiation. With the higher HIV-associated mortality rate at their clinic, however, the threshold value for IRIS-associated mortality would almost certainly be higher than 4.6%. This example demonstrates the need to address the issue of optimal timing of ART initiation in HIV-infected patients with TB using data entirely from a developing world setting.
A randomized clinical trial would be the best way to address this issue, but as Lawn and Wood mention, the appropriate clinical trials have been slow to enroll. Therefore, a formal decision analysis with data from the developing world would be beneficial to confirm that early ART initiation is indeed preferred in this setting. We have recently shown that the conclusions of decision analyses pertaining to infectious disease topics usually correspond with the results of subsequent clinical studies that address the same scientific question.5 This provides support for the validity of decision analysis in guiding clinical decision making.
It should be noted that for any future decision analysis on this topic, sufficient data are necessary for variables that are important but for which current data are limited, including HIV-related mortality, drug toxicity and associated mortality, and TB failure and relapse-all for persons who initiate ART within 2 months of anti-TB therapy versus those who initiate after 2 to 6 months of therapy.
Joshua T. Schiffer, MD*†
Timothy R. Sterling, MD‡§
*Division of Infectious Diseases Department of Medicine University of Washington Seattle, WA
†Department of Medicine Johns Hopkins University Baltimore, MD
‡Division of Infectious Diseases Department of Medicine Vanderbilt University Medical Center Nashville, TN
§Center for Health Services Research Department of Medicine Vanderbilt University Medical Center Nashville, TN
1. Schiffer JT, Sterling TR. Timing of antiretroviral therapy initiation in tuberculosis patients with AIDS: a decision analysis. J Acquir Immune Defic Syndr
2. Burman W, Vernon A, Khan A, et al. Timing of antiretroviral therapy during treatment of HIV-related tuberculosis [abstract 846]. Presented at: 43rd Annual Meeting of the Infectious Diseases Society of America; San Francisco, CA; 2005.
3. Lawn SD, Myer L, Bekker LG, et al. Tuberculosis-associated immune reconstitution disease: incidence, risk factors and impact in an antiretroviral treatment service in South Africa. AIDS
4. Lawn SD, Bekker LG, Miller RF. Immune reconstitution disease associated with mycobacterial infections in HIV-infected individuals receiving antiretrovirals. Lancet Infect Dis
5. Bress JN, Hulgan T, Lyon JA, et al. Agreement of decision analyses and subsequent clinical studies in infectious diseases. Am J Med